Russ Cox on #Mastodon about simple and fast #floating-point printing and parsing in #Golang #toread #small-is-beautiful
on 02026-01-25#PDF #paper “#ACL2 Support for #Floating-point Computations” #formal-methods
on 02025-11-28“"Flocq" (Floats for Coq) is a formalization of #floating-point arithmetic for the #Coq proof assistant. It provides a comprehensive library of theorems on a multi-radix multi-precision arithmetic. It also supports efficient numerical computations inside Coq.” #formal-methods
on 02025-11-28#video on #videogames #floating-point bugs in timing #toread
on 02025-10-22#moon-child and #rygorous discussing #GPU #floating-point and #FMA and dot products and round-to-negative vs. round-toward-zero etc.
on 02023-10-07#PDF #paper by #Kahan about #floating-point being compromised in the pursuit of #performance to cheat on SPEC benchmarks
on 02023-10-07Looks like the #T-Head #RISC-V #hardware fails to implement #floating-point underflow correctly, at least the C910, C920, and probably C906, and in particular the LicheePi 4A.
on 02023-09-05Berkeley SoftFloat is a high-quality implementation of #floating-point math (not by Monte Davidoff).
on 02019-02-14#Apple II ROM information from 1997/2004; includes a hex dump of the 12KiB of the ROM, the SWEET_16 source code and documentation, the #floating-point package source code, etc. #history #pdf
on 02018-10-13#PDF on #FFT #algorithms on Intel GPUs, including some architectural details: each EU has 7 hardware threads, each with 128 32-byte registers, 512 bytes per work item in SIMD-8 mode, but only 2 32-bit #floating-point ALUs #GPGPU
on 02018-10-05#IEEE-754 #floating-point manages to make even the simplest functions into complicated quagmires! #Python
on 02016-07-05